Cloudflare services offline due to data centre power outages

Cloudflare services offline due to data centre power outages

Problems with power to multiple Cloudflare data centres are having a significant impact on the company’s services. Among other things, customers cannot access their Cloudflare dashboard and related APIs for management and configuration functionality.

The problems in the data centres are related to power delivery, according to Cloudflare. Power problems in the U.S. state of Oregon have affected several data centres, Cloudflare said in an initial update. With several generators also not working, these went offline.

Services affected by the outage include the Cloudflare dashboard, the Cloudflare AOI, Logpush, WARP/ Zero Trust device posture, Stream API, Workers API and the Alert Notification system. The status of these can be seen on this website.

The cached file delivery via Cloudflare CDN and Cloudflare Edge security functionality is not suffering the same outage problems. Also, the flow of network traffic throughout the Cloudflare network is not experiencing any problems.

Service restoration started

Cloudflare is now working with its data centre providers to restore service with all hands. In the latest update on the incident, the application and network performance specialist says it has switched to its disaster recovery facility and is working to bring the infrastructure back online.

The incident may have lost so-called Logpush logs and analytics data, however. The total damage of this cannot be determined at this time, according to the company. However, logs created at the end of the incident may be recoverable.

Previous incident

The problems with the power supply in Cloudflare’s data centres is the second incident in a week that the application and network performance specialist is experiencing. Earlier this week, several other services also went offline. These included Cloudflare Sites and Services, which include Access, CDN Cache Purge, Dashboard, Images, Pages, Turnstile, Waiting Room, WARP and Workers KV.

Investigation revealed that this outage was caused by a misconfiguration in the tool used to deploy a new Workers KV build.

Also read: Cloudflare signals DDoS record: 201 million requests per second